In the proposed study we plan to examine the involvement of plasminogen activator in the process of malignant dissemination and to try to identify factors which might be responsible for a change in malignant phenotype of tumor cells. The role of plasminogen activator will be tested in vivo in two different experimental systems--the chick embryo and the nude mouse. Human carcinoma (HEp3), which produces large amounts of urokinase type plasminogen activator, grows and rapidly and predictably metastasizes into the organs of the chick embryo. We have shown that anti-plasminogen activator antibody substantially reduces, and often prevents, metastasis of this tumor. To determine in which step of tumor dissemination plasminogen activator might play a role, we will continue using the chicken embryo system. In this system, the metastasis is rapid and the disseminating cells are accessible for analysis at any given time. The study will be conducted using histological screening of sections of control and antibody treated tumor-bearing embryos and also, by tracing and spread of radioactively labelled tumor cells. To test the generality of our findings, similar experiments will be performed in another host--the nude mouse, in which this tumor grows and metastasizes. Tumor cells can be found invading locally into the neighboring tissues, the lymphatics, blood vessels and distally to lungs. We plan to examine the effect of antibodies (poly and monoclonals) on the local invasion and metastasis. Another highly metastatic tumor--a human melanoma--which produces exclusively tissue type plasminogen activator will be examined in the same context. The second part of the project will involve the factors which affect malignant phenotype of the HEp3 tumor. We have found that the malignant phenotype is maintained only in vivo and that during growth in cell culture it is progressively reduced to the point of extinction. This is a reversible change; malignancy reappears with prolonged exposure to in vivo conditions. By finding which properties are associated with the malignant and the nonmalignant phenotypes and how the conversion in this system can be affected, we hope to gain some information on the control of expression of malignancy in general. (B)